The deposition of excess collagen in the myocardial extracellular matrix contributes to increased myocardial stiffness leading to diastolic (D) dysfunction and heart failure (HF), as well as increases in brain natriuretic peptide (BNP) which are predictive of outcome. In the elderly, DHF is at least as common as systolic(S) HF and both have poor outcome. Amino-terminal propeptide of type Ill procollagen (PIIINP) and carboxy-terminal propeptide of type I procollagen (PIP) are serum markers of fibrosis. As an ancillary study within the Cardiovascular Health Study (CHS), a prospective epidemiologic study of 5,888 community-dwelling elderly participants who to date have had 12 years of follow-up, we will test hypotheses that: 1. myocardial fibrosis is increased in prevalent DHF as well as SHF, compared to controls; 2. myocardial fibrosis increases with aging and is a predictor of incident DHF and SHF; 3. myocardial fibrosis is a predictor of outcome - total and cardiovascular death (including sudden death), non-fatal myocardial infarction and stroke-in subjects with prevalent HF, and 4. BNP serum levels are associated with myocardial fibrosis in elderly with and without HF. In a nested case control design we will use stored frozen serum to measure PIIINP, PIP, and BNP in 175 participants with DHF, 146 with SHF, and 679 age and gender matched controls including 339 healthy normals. Two research echocardiograms (echo) done in years bracketing the blood measures, as well as clinical echo reports at the point of care will be utilized to assess cardiac structural and functional covariates of prevalent HF, those preceeding incident HF, and to characterize DHF (HF with normal ejection fraction-EF-), SHF (HF with low EF), and asymptomatic systolic ysfunction (low EF but no HF). The large number of clinical assessments, functional measures, blood assessments (e.g. chemistries, inflammation factors, lipids) and measures of subclinical disease (e.g. carotid intima-medial thickness, ankle-arm index) within the CHS will permit analyses determining the association of myocardial fibrosis with co-morbidities, the assessment of myocardial fibrosis in healthy vs. normative aging, and the influence of covariates on relationships between fibrosis, HF and outcome. In addition to improving prediction of outcome of HF in elderly, the findings of this research may aid in designing clinical trials of treatment and prevention targeted to myocardial fibrosis.